modulo one with tolerance r = SW_CMOD(r, tol) It calculates modulo one with tolerance, numbers larger than 1-epsilon > 1-tol will be converted to -epsilon. See also MOD.